U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number:  FHWA-HRT-12-054    Date:  December 2012
Publication Number: FHWA-HRT-12-054
Date: December 2012


Methodologies to Measure and Quantify Transportation Management Center Benefits: Final Synthesis Report

4. Spatial and Temporal Data Structures

4.1 Data Capabilities of FMSs and Traffic Signal Systems

The following list describes a set of data collection, storage, and data manipulation capabilities that are common to most FMSs:

Time periods for data collection and archiving that are commonly employed by FMS are shown in table 8 .

Table 8. Data periods.

Data Period Description Typical Period Examples of Use
Discrete data element Each event Crash report, incident report, and equipment event or failure
Data sampling or collection period 20 s to
1 min
Traffic detector collection period for field detectors
Action periods 1 to
10 min
Data accumulation periods for TMC actions such as traffic map displays, data filter updates, system-wide ramp metering, incident management, automatic dynamic message sign (DMS) messaging, and system tuning
Common reporting and analysis interval 5 min,
15 min, 1 h
Studies of traffic patterns by TMC personnel and others
Daily reports 1 day Daily data consolidations and planning
Annual reports 1 year Performance evaluations and planning

An example of the general relationship between data uses and data characteristics is shown in table 9 .(12)

Table 9. Data uses and characteristics.

Type of Data Use User Data Used Source
Long-term planning PSRC AADT volume Washington State Department of Transportation (WSDOT) Annual Traffic Report(13)
Highway Performance Monitoring System VMT WSDOT Data Office
24-h and peak volume counts Ramp & Roadway Report(14)
24-h volume counts City and County Tube Collections
WSDOT Planning Office Volume counts Annual Traffic Report(13)
Forecasted efficiency data PSRC
Performance monitoring PSRC AADT volume Annual Traffic Report(13)
24-h and peak volume counts Ramp & Roadway
24-h volume counts City and County Tube Collections
Long-range planning and project planning WSDOT Transportation Data Office AADT volumes Annual Traffic Report(13)
Projected volume data PSRC
Turning movements Northwest Region Planning Office
Vehicle occupancy Northwest Region Planning Office
Vehicle classification Northwest Region Planning Office
Specific volume counts Northwest Region Planning Office
Travel time and speed Consultants
Transit use Consultant
Pedestrian and bicycle counts Consultants
Performance Monitoring WSDOT Office Of Urban Mobility Volume counts Travel-Time Reporting and Integrated Performance System (TRIPS)
Incident data TRIPS
Research Washington State Transportation Center, Transportation Northwest, and the University of Washington Researchers 20 s, 1 min, 5 min, 15 min WSDOT Transportation System Management Center
Volume counts and lane occupancy Ramp & Roadway Report(14)
Peak volume counts Annual Traffic Report(13)
AADT volumes Automated data collection stations (ADCSs), autoscope
Speed WSDOT Data Office
Vehicle classification ADCS, autoscope
Vehicle occupancy Washington State Transportation Center

Note: PRSC denotes the region's Metropolitan Planning Organization.

This project develops methodologies for employing FMS data to generate many of the evaluation measures described in table 6. Data collected every 5 min are the building blocks for freeway-based measures that develop or utilize travel time or delay. Figure 2 shows an example of a data aggregation structure for freeway point detector data.(15)

Figure 2. Illustration. Example of data aggregation structure. This figure shows the progression of the data aggregation on a generic six-lane divided highway segment (three westbound lanes and three eastbound lanes) through four different progressions. The first progression is labeled lane-by-lane level in which individual traffic sensors are shown in each lane at 0.5-mi spacings (24 small grey squares with 4 sets of 3 (1 per lane) stacked vertically eastbound and westbound). Traffic sensors collect data in each lane at 0.5-mi nominal spacing with arrows pointing to two sensors (one westbound and one eastbound). An arrow points down toward the next progression for this same highway segment. The second progression is labeled station level. Rectangular grey boxes extend across the groups of three individual lane detectors from the first progression. Summary statistics are computed across all lanes in a given direction with arrows pointing to two of the rectangles detector groups (one westbound and one eastbound). An arrow points down toward the next progression for this same highway segment. The third progression is labeled link level. Eight larger grey rectangles are shown on the highway segment (four westbound and four eastbound) arranged horizontally across all three lanes. The rectangles are labeled link travel time and vehicle miles of travel. Point-based properties extrapolated to roadway links 0.5 to 3 mi in length, with arrows pointing to two of the rectangles (one eastbound and one westbound). An arrow points down toward the next progression for this same highway segment. The fourth progression is labeled section level. The four individual link rectangles in each direction in the third progression have been replaced with one large grey rectangle for a total of two large grey rectangles (one east bound and one westbound) across all the links in each direction. The rectangles are labeled directional roadway section travel time and vehicle miles of travel. Link properties are summed to analysis sections 5 to 10 mi in length with arrows pointing to the two rectangles.

Figure 2. Illustration. Example of data aggregation structure.

Although the capability exists in traffic signal systems to collect and archive volume, occupancy, and speed data (at a particular location), other than some adaptive signal systems, traffic signal systems generally do not have the capability to provide data for the measures needed to obtain key parameters such as travel time and delay. Section 6 of this report describes some recently developed techniques that may be employed to provide these measures. To be consistent with independent volume measures such as automatic traffic recorders and manual count collections, a 15-min period is recommended as the basic surface street evaluation interval.

4.2 Spatial Requirements and Data Structures for Evaluation

A data structure concept is required to relate the data sources (e.g., detector data, crash reports, and incidents) to a construct that may be used for evaluation purposes. An example of a construct that might be used for evaluation purposes includes the following:

FMSs generally contain a software capability to provide a reference framework to relate detectors to the link structure for the freeway network. If the FMS does not have such a capability, the evaluation methodology must provide it. A reference system that is based on traffic flow entry and exit points is preferred for the following reasons:

An example of a reference system that meets this requirement is shown in figure 3 .

A link represents a section of the mainline between vehicle access or egress points. The concept of a domain is introduced in figure 3 to relate data from freeway surveillance stations to mainline links. Domains relate links and DMSs [2] to the roadway locations receiving information from a particular point detector station. As shown in the figure, each domain is related to a particular detector station. Domain boundaries are established at link nodes and at the DMS. Where a link encompasses more than one detector station, domain boundaries are used to separate the regions for which each detector station will be employed. Note that none of the detectors in figure 3 exist within the physical boundaries of domain 4; that domain obtains its information from detector station 4. Section 6 of this report discusses detector deployment requirements.

Figure 4 shows a similar diagram for probe-based surveillance. The asterisks identify locations for probe travel time measurements. These boundaries may be established by physical equipment locations (i.e., toll tag reader locations or locations of Bluetooth® readers) or may be virtual boundaries for other types of probe detection systems such as those based on a Global Positioning System (GPS). While it is sometimes possible to co-locate virtual or actual boundaries with link boundaries, this is not always the case. The probe-measured travel times are converted to speeds, and these speeds, in conjunction with link lengths, are used to estimate travel link travel times. Probe-based detection does not provide volume estimates, so supplementing these data with other information is required for the system-based measures required for benefit-cost analysis. In order to obtain system-wide delay and travel time measures with probe detection, at least one source of volume per link is required. Technologies for implementing probes and other sensors are discussed in section 6 of this report.

Figure 3. Illustration. Example of link, domain, and detector station relationships. This figure is shows the complex relationship between the links, detector stations, and domains. Lines and symbols are used to represent five links (labeled L1 through L5) and their boundaries defined by travel nodes, with six detector stations shown as black squares (labeled K1 through K6) spread out across seven domains (labeled D1 through D7). Links may cross several different domain boundaries and receive information from one or more detectors. The figure also shows that some detectors may provide data to more than one domain.

Figure 3. Illustration. Example of link, domain, and detector station relationships.

Figure 4. Illustration. Example of link, domain, and probe site relationships. This figure shows the complex relationship between links, domains, and probe sites. Lines, arrows, circles, and asterisks are used to represent links and their boundaries, probe sensing regions, and domains. Domains may gather information from probe sensing regions that stretch across multiple links. A single probe sensing region may be the only information source for multiple domains.

Figure 4. Illustration. Example of link, domain, and probe site relationships.

4.3 Temporal Relationships

For archiving purposes, FMS volume, speed, and occupancy data from point detectors may be stored at 5-min intervals and aggregated into 15-min and 1-h intervals, as in the Florida STEWARD system.(8) The 5- and 15-min intervals provide convenient processing intervals for many of the delay-related computations described in section 5 of this report. Building on these concepts, a useful methodology develops these measures using the spatial/temporal relationship shown in figure 5. The methodology described uses the domain concept as the basis for freeway mainline data accumulation (see figure 3 and figure 4).

Figure 5. Illustration. Data accumulation methodology. This illustration shows the data accumulation methodology. Spatial relationships are on the left-hand column and include domain, link, route, and system. Along the top row, the time intervals are listed, which include5 min, 15 min, 1 h, 1 day, and 1 year. At the 5-min interval, a solid black arrow points down from the domain row to the link row. A dashed black arrow continues down through the route row to the system row. A solid black arrow points to the right in the link row from the 5-min column to the 15-min column. A dashed black arrow continues from the 15-min column through the 1-h and 1-day columns to the 1-year column in each of the link, route, and system rows. A dashed black arrow in the 15-min column points down from the link row through the route row to the system row.

Figure 5. Illustration. Data accumulation methodology.

Detector data are used to obtain these measures at the domain level for 5-min periods and are accumulated at the link level. The 15-min period at the link level is a convenient building block for many of the evaluation measures. The path to computing this level for the 15-min period is shown by the solid trace. The dashed traces show the paths to other spatial levels and time periods. Depending on the particular measure to be computed and the purpose (reports, etc.), the 15-min data may be aggregated by time according to the particular spatial relationship required for the purpose.

2 Although not strictly needed for the detector to link relationships, figure 3 includes DMS in the domain definitions to facilitate the implementation of messaging using a common reference frame.


Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101